reviewed 17.10.2017 accepted 08.11.2017 a spatio-temporal

© Polish Academy of Sciences (PAN) in Warsaw, 2018; © Institute of Technology and Life Sciences (ITP) in Falenty, 2018

© Polish Academy of Sciences (PAN), Committee on Agronomic Sciences JOURNAL OF WATER AND LAND DEVELOPMENT Section of Land Reclamation and Environmental Engineering in Agriculture, 2018 2018, No. 36 (I–III): 183–197 © Institute of Technology and Life Sciences (ITP), 2018 PL ISSN 1429–7426 Available (PDF): http://www.itp.edu.pl/wydawnictwo/journal; http://www.degruyter.com/view/j/jwld

Received 05.05.2017 Reviewed 17.10.2017 Accepted 08.11.2017 A – study design B – data collection C – statistical analysis D – data interpretation E – manuscript preparation F – literature search

Spatio-temporal variation of rainfall over Bihar State, India

Pratibha WARWADE1) ABCDEF, Shalini TIWARI1) ABCDEF,

Sunil RANJAN1) ABCDEF, Surendra K. CHANDNIHA2) ABCDEF,

Jan ADAMOWSKI3) ABCDEF

1) Central University of Jharkhand, Centre for Water Engineering and Management, Ranchi, India 2) National Institute of Hydrology, Surface Water Hydrology Division, Roorkee, Uttarakhand, India 3) McGill University, Department of Bioresource Engineering, 21 111 Lakeshore Road, Ste Anne de Bellevue, QC H9X 3V9,

Canada; [email protected]

For citation: Warwade P., Tiwari S., Ranjan S., Chandniha S.K., Adamowski J. 2018. Spatio-temporal variation of rainfall over Bihar State, India. Journal of Water and Land Development. No. 36 p. 183–197. DOI: 10.2478/jwld-2018-0018.

Abstract

This study detected, for the first time, the long term annual and seasonal rainfall trends over Bihar state, In-dia, between 1901 and 2002. The shift change point was identified with the cumulative deviation test (cumula-tive sum – CUSUM), and linear regression. After the shift change point was detected, the time series was sub-divided into two groups: before and after the change point. Arc-Map 10.3 was used to evaluate the spatial distri-bution of the trends. It was found that annual and monsoon rainfall trends decreased significantly; no significant trends were observed in pre-monsoon, monsoon, post-monsoon and winter rainfall. The average decline in rain-fall rate was –2.17 mmꞏyear–1 and –2.13 mmꞏyear–1 for the annual and monsoon periods. The probable change point was 1956. The number of negative extreme events were higher in the later period (1957–2002) than the earlier period (1901–1956).

Key words: climate change, hydrology, rainfall, spatio-temporal variation

INTRODUCTION

The spatial and temporal distribution of rainfall is important in hydrology, water resource management, agriculture, as well as policy making and planning with regards to climate change [HAMMOURI et al. 2015]. The uneven distribution of rainfall may result in a mismatch between water availability and demand in which case irrigation structures are required to re-distribute water in accordance with the requirements of a specific region. The incidence of extreme rainfall events is also an important parameter for the design of dams, spillways, flood protection structures and drainage networks [JAGADEESH, ANUPAMA 2014]. The Intergovernmental Panel on Climate Change

(IPCC) reported that the intensity of spatial, inter-seasonal and inter-annual variability of rainfall events have increased in Asia over the past few decades [TEAM et al. 2014. A rise in mean global temperature of about 0.74℃ between 1906 and 2005 and greater rainfall variability have also occurred [SOLOMON et al. 2007; ŻMUDZKA 2012]. These changes affect vari-ous issues such as crop water requirements, evapo-transpiration, ground water level depletion, environ-mental flows, trans-boundary issues and interstate disputes, among others. Many hydrological studies have been carried out around the world with regards to water issues related to rainfall and temperature (e.g. NICHOLLS, LAVERY [1992], RODRIGUEZ-PUEBLA et al. [1998], MOHAPATRA et al. [2003], GOSWAMI et al.

DOI: 10.2478/jwld-2018-0018

184 P. WARWADE, S. TIWARI, S. RANJAN, S.K. CHANDNIHA, J. ADAMOWSKI

© PAN in Warsaw, 2018; © ITP in Falenty, 2018; Journal of Water and Land Development. No. 36 (I–III)

[2006], GALLANT et al. [2007], GUHATHAKURTA, RA-JEEVAN [2008], TASCHETTO, ENGLAND [2009], CHOW-DHURY, BEECHAM [2010], PHILANDRAS et al. [2011], AYALEW et al. [2012], MARTINO et al. [2012], CZAR-NECKA, NIDZGORSKA-LENCEWICZ [2012], PETRIE et al. [2014], ARAGHI et al. [2015], CHANDNIHA et al. [2016], WARWADE et al. [2016], INAM et al. [2017]).

Studies from around the world show that rainfall variability is increasing on both a spatial and a tempo-ral basis. For example, in the Australian and Canadian prairies [AKINREMI et al. 2001]; in the Haihe River Basin, China [LU et al. 2015]; and in Turkey, particu-larly in January, February and September [PARTAL, KAHYA 2006]. Various studies on rainfall trends have also been carried out in in India [GUHATHAKURTA, RAJEEVAN 2008; KOTHYARI, SINGH 1996; PAR-

THASARATHY, DHAR 1974; 1978; SARKAR et al. 2004; SEN ROY, BALLING 2004], and it has been reported that Indian monsoon rainfall is strongly affected by atmospheric factors viz. regional pressure anomalies [DUGAM, KAKADE 2003], cyclonal activities [PAT-

TANAIK, RAJEEVAN 2007] and sea surface temperature variations [GOSWAMI et al. 2006], thereby making monsoon rainfall less predictable [MANI et al. 2009]. PATTANAIK, RAJEEVAN [2010] assessed long term trends in the Indian monsoon season from 1951 to 2005, and found extreme rainfall events increased and low rainfall events decreased. Extreme climatic and weather events have increased over the last few dec-ades [KARL, EASTERLING 1999]. In Kerala in South India, a significant decreasing rainfall trend was found in the southwest monsoon and an increasing trend was found in the post-monsoon season [KRISH-

NAKUMAR et al. 2009]. In various Indian states such as Punjab, Haryana, Rajasthan and Madhya Pradesh, increasing rainfall trends in the annual and southwest monsoon occurred during 1901 to 1982 [PANT, HINGANE 1988]. In central India and the adjoining part of the Peninsula, a trend in rising annual rainfall between 1901 and 1960 was recorded [PARTHASARA-

THY, DHAR 1974]. Although various studies on climate change have

been carried out in the central, northern and southern parts of India, little has been done in the eastern re-gion. Our investigation therefore focuses on Bihar, an eastern state of India and studies the annual and sea-sonal (pre-monsoon, monsoon, post-monsoon and winter) rainfall variation in terms of trend, magnitude and extreme events at 37 weather stations. Various statistical techniques such as the Mann–Kendall (MK) test, Sen’s slope estimator, change percentage and break point analysis are used and supported by a linear regression test. This rainfall trend analysis is important to facilitate the planning and management of sustainable water resources in the region.

STUDY AREA

The state of Bihar borders on Uttar Pradesh (to the West), Jharkhand (to the South) and West Bengal

(to the North-East) and is the 13th-largest state in In-dia with a surface area of 94,163 km2 (Fig. 1). It is the third most populous state with over 100 million in-habitants. About 6,764 km2 are covered by forest which is about 7.2% of the total geographical area of Bihar state. The state is bifurcated by the Ganges Riv-er which flows from West to East. Bihar lies between 24°37'58" N to 27°33'86" N latitude and between 83°21'47" E to 88°19'41" E longitude. The average temperature varies from 10°C in the winter to 34°C in the summer and the weighted average annual rainfall of the state is approximately 1170 mm [India Water Portal 2010].

Fig. 1. Geographical locations of different districts over the study area (Bihar State); source: own elaboration

MATERIALS AND METHODS

The long term (1901–2002) monthly rainfall data were acquired from the India water portal site [India Water Portal 2010]. This monthly rainfall time series has been used in other studies [CHANDNIHA et al. 2016; DUHAN, PANDEY 2013]. In the present study, the 1901 to 2002 rainfall time series were segregated into 5 data sets viz. annual (1) and seasonal (4) as characterized by the India Meteorological Department (IMD) guidelines: monsoon (June–Sept); post-mon-soon (Oct–Nov); winter (Dec–Feb) and pre-monsoon (Mar–May). Missing values and consistency of the long-term data set were checked and the presence of serial dependency over the time series was determined by applying lag-1 auto-correlation [CUNDERLIK, BURN 2004]. The non-parametric Mann–Kendall trend and Sen’s slope estimator test were applied to monthly, annual and seasonal data sets for all the dis-tricts in Bihar [KENDALL 1975; PINGALE et al. 2016; SEN 1968; THEIL 1950]. The Mann–Kendall test was applied at a 95% confidence level and overall results are represented by positive significant, negative signifi-cant, positive and negative trends. Regression analysis was also utilized for estimating the relationship among the variables [JAGADEESH, ANUPAMA 2014]. A possible shift change (break point) was detected using the cumulative deviation test at a confidence

Spatio-temporal variation of rainfall over Bihar State, India 185


level of 95% [BUISHAND 1982]. This test can still ap-ply when there are slight departures from normality. The cumulative deviation (CD) test is based on the adjusted partial sums or cumulative deviations from the mean. Sequential analysis, specifically the cumu-lative sum (CUSUM) test, was used to monitor the change detection in the series [TABARI et al. 2012].

INVERSE DISTANCE WEIGHTED TECHNIQUE

In the present study, the inverse distance weighted (IDW) interpolation technique was used for the spatial maps of precipitation for 37 stations in Bi-har State [LEBEL et al. 1987; SINGH, CHOWDHURY 1986].

The inverse distance weighting tool is already available in ArcMap. In the IDW methodology, the weight of any known point is set to be inversely pro-portional to its distance from the estimated point [CHEN, LIU 2012]. It is calculated as follows:

n

i i

n

ii

i

d

xd

x

1

1

1

1

ˆ

(1)

Where: x̂ = value to be estimated; xi = known value; d1, d2, d3, …, dn = distance from the n data points to the point estimated n.

BRIEF METHODOLOGY OF MANN-KENDALL TEST AND SEN’S SLOPE ESTIMATOR

1. The Pettitt Mann–Whitney (PMW) test distin-guishes the most likely change year in the yearly precipitation sequence [PETTIT 1970]. The details of the PMW test is given below [DUHAN, PANDEY 2013]: Assume a time series {X1, X2, …, Xn} with a length n and where t is the time of the most like-ly change point. Two samples, {X1, X2, …, Xn} and {Xt+1, Xt+2, …, Xn}, can then be derived by divid-ing the time series at time t. An index, Ut, is de-rived via [DUHAN, PANDEY 2013]:

Ut=∑ ∑ sgn Xi –XjTj=1

ti=1 (2)

sgn Xi –Xj =

1 if Xi –Xj >0

0 if Xi –Xj =0

–1 if Xi –Xj <0

(3)

A plot of Ut against t for a time series with no change point would result in a continually increas-ing value of Ut. However, if there is a change point (even a local change point) then Ut would increase up to the change point and then begin to decrease. This increase followed by a decrease may occur several times in a time series, indicat-ing several local change points. So there is still the question of determining the most significant

change point. We can identify the most significant change point t where the value of |Ut| is maximum [DUHAN, PANDEY 2013].

Kt=max1≤t≤T

|Ut|

The approximated significance probability p(t) for a change point [PETTITT 1979] is given as:

23

26exp1

TT

Kp T (4)

The change point is statistically significant at the time t with a significance level of α when proba-bility p(t) exceeds 1 – α. The autocorrelation coefficient distinguishes the nearness of serial correlation within the data series [SIEGEL, CASTELLAN 1988].

2. The Mann–Kendall (MK test has been applied to the non-auto correlated/auto correlated sequences to distinguish the presence of patterns in yearly and seasonal precipitation [HAMED, RAO 1998; KENDALL 1975]. The MK statistic, S, is defined as [HAMED, RAO 1998; KENDALL 1975]:

1n

1i

n

1ijij xxsgnS

(5)

Where: x1, x2, x3, …, xn represents n data points where xj represents the data point at the time j of data (time series); and xj – xi = θ

01

00

01

if

if

if

Sgn (6)

Under the assumption that the data are independ-ent and identically distributed, the mean and vari-ance of the S statistic in Eq. (2) are given by [KENDALL 1975] as:

E[S] = 0

18

5t21tt5n21nnSVar

m

1iiii

(7)

Where: m = the number of groups of tied ranks, each with ti tied observations.

The original MK statistic, designated by Z and the corresponding p-value (p) of the one-tailed test was computed as [DUHAN, PANDEY 2013]:

0SIfSVar

1S

0SIf0

0SIfSVar

1S

Z

(8)

Z5.0p



Z

0

2

t

dte2

1Z

2

(9)

If the p-value is small enough, the trend is quite unlikely to be caused by random sampling. The Z values are approximately normally distributed, and a positive Z value larger than 1.96 (based on nor-mal probability tables) denotes a significant in-creasing trend at the significance level of 0.05, whereas a negative Z value lower than –1.96 shows a significant decreasing trend. Thiel and Sen's slant estimator test served to esti-mate the magnitude of a trend [SEN 1968; THIEL 1950].

3. In this method, the slope estimates of N pairs of data are first calculated using the following ex-pression [SEN 1968]:

n,......,3,2,1iforkj

xxQ kj

i

(10)

Where: xj and xk are data values at time j and k (j > k) respectively. The median of these N values of Qi is Sen’s esti-mator of slope, which is calculated as [SEN 1968]:

evenisNQQ

oddisNQ

N

NN

22

22/1

2

1 (11)

A positive value of β indicates an upward (increas-ing) trend and a negative value indicates a downward (decreasing) trend in the time series data.

PERCENTAGE CHANGE

The percentage change phenomenon is used for comparison between two scenarios in a time series. Change percentage was computed by approximating it with a linear trend [DUHAN, PANDEY 2013; PINGALE et al. 2015; YUE, HASHINO 2003]. The percentage change was calculated as:

100%

Mean

yearofLengthChangePercentage

RESULTS AND DISCUSSION

SERIAL CORRELATION IN THE SERIES

The presence of any positive or negative correla-tions may affect the trend detection procedure [YUE, HASHINO 2003]. Hence, lag-1 serial correlation is used to check the presence of any existing trends in the series before applying the MK test. Results of the autocorrelation test show that all the series were found to be non-autocorrelated (Tab. 1). The MK test was then applied to the non-autocorrelated data series at the 5% significance level over 102 years. Figure 2 shows the deviation of rainfall from the mean value with the help of linear regression. The annual and

monsoon graphs illustrate prominent decreasing rain-fall trends and a sudden change after the mid-century. Pre-monsoon, post-monsoon and winter seasons do not show clear trends.

Table 1. Values of change points and autocorrelation coef-ficients

Sta-tion No

Station

Cumula-tive de-viation

(t)

CUSUM (t)

Autocorrelation coefficient

(ρk)

1 Araria 1956 1959 0.141 2 Aurangabad 1978 1950 0.221 3 Banka 1959 1973 0.101 4 Begusarai 1956 1963 0.170 5 Bhagalpur 1956 1959 0.096 6 Bojpur 1964 1964 0.279 7 Buxar 1964 1973 0.264 8 Darbhanga 1956 1974 0.159 9 Gaya 1978 1950 0.224 10 Gopalganj 1987 1987 0.308 11 Jamui 1978 1963 0.167 12 Jehanabad 1978 1964 0.255 13 Bhabua 1982 1973 0.155 14 Katihar 1956 1959 0.062 15 Khagaria 1956 1959 0.138 16 Kishanganj 1956 1956 0.158 17 Lakhisarai 1959 1963 0.186 18 Madhepura 1956 1959 0.123 19 Madhubani 1956 1959 0.156 20 Munger 1959 1973 0.153 21 Muzafferpur 1987 1974 0.220 22 Nalanda 1964 1964 0.224 23 Nawada 1978 1964 0.219 24 Paschim Champaran 1963 1988 0.256 25 Patna 1964 1978 0.247 26 Purba Champaran 1987 1987 0.265 27 Purnia 1956 1959 0.101 28 Rohtas 1978 1973 0.197 29 Saharsa 1956 1959 0.127 30 Samastipur 1956 1959 0.165 31 Saran 1987 1974 0.287 32 Sheikhpura 1959 1964 0.203 33 Sheohar 1987 1987 0.229 34 Sitamarhi 1956 1987 0.198 35 Siwan 1987 1987 0.307 36 Supaul 1956 1959 0.135 37 Vaishali 1964 1978 0.226

Source: own study.

SHIFT IN RAINFALL TIME SERIES FROM 1901 TO 2002

Table 1 indicates 1956 as the most probable change point in the annual rainfall time series as ob-tained from the CD test and supported by the CUSUM test and linear regression (Fig. 2). Hence the entire time series (1901–2002) was divided into two seg-ments, from 1901 to 1956 (before change point) and from 1957 to 2002 (after change point), and further climate variability was analysed considering the di-vided time series.



Fig. 2. Annual and seasonal rainfall trend (1901–2002) deviation from the mean value by the linear regression method; source: own study

Fig. 3. Rainfall variability on an annual and seasonal basis at all stations; source: own study

RAINFALL VARIABILITY

Predicting variation in rainfall is important from an agricultural point of view for accurate estimation of crop water requirements. Rainfall variability pat-terns over 1901 to 2002 for 37 stations in Bihar using the coefficient of variation (CV) indicates that the in-ter-annual variability of post-monsoon rainfall is great-er than that of the annual rainfall. The CV ranges from 14.37% to 21.1%, from 16.38% to 22.81%, from 53.99% to 67.28%, from 38.44% to 71.88 % and from

67.89% to 78.08% for annual, monsoon, winter, pre-monsoon and post-monsoon, respectively. The aver-age rainfall in Bihar (with average variability) is 1169.24 mm (17.69%), 988.05 mm (19.19%), 65.54 mm (57.02%), 79.04 mm (72.50%) and 38.97 mm (60.04%) in annual, monsoon, pre-monsoon, post- -monsoon and winter, respectively (Fig. 3). Overall, there is a high variation in rainfall; areas with high inter-annual variability in rainfall are generally more susceptible to floods and droughts [PANDEY, RAMA-

SATRI 2001; TURKEŞ 1996].

Rai

nfal

l ano

mal

y, m

m

Rai

nfal

l ano

mal

y, m

m

Rai

nfal

l ano

mal

y, m

m

Rai

nfal

l ano

mal

y, m

m

Rai

nfal

l ano

mal

y, m

m

Coe

ffic

ient

of v

aria

tion,

%



The highest variability was observed in post- -monsoon precipitation (78.08%) and the lowest vari-ability in annual precipitation (14.37%) over the 102 years as reported by CHANDNIHA et al. [2016] for Jharkhand, a border state.

DISTRIBUTION OF AVERAGE RAINFALL IN BIHAR

The average rainfall distribution of the monsoon season over 102 years as a proportion of average an-nual rainfall for 37 stations (Fig. 4) shows that the monsoon contributed an average of 84.58% to the total average annual rainfall in Bihar. Individual sta-tions experienced not less than 73.49% of their rain-fall during the monsoon. The other seasons did not make a significant contribution to the average annual rainfall. Hence, further trend analysis was carried out for annual and monsoon rainfall time series.

TRENDS IN RAINFALL

Spatial and temporal trends in annual and seasonal rainfall between 1901 and 2002

The direction (Z value), magnitude of the trend (Q), and percentage change (change %) for annual and seasonal rainfall for the entire state of Bihar over 102 years (1901–2002), 56 years (1901–1956) and 46 years (1957–2002) are shown in Table 2 (Table 2 is on a sea-sonal basis, while Table 3 is on a station basis).

Positive and negative values of Z indicate increas-ing and decreasing trends, respectively. Significant negative trends for annual and monsoon rainfall for 1901–2002 and 1957–2002 and a significant positive trend for 1901–1956 were observed. The rest of the seasons showed non-significant trends for all time periods. Annual and monsoon rainfall declined at rates of –2.17 mmꞏyear–1 and –2.12 mmꞏyear–1 during 1901 to 2002, respectively, and at much greater rates from 1957 to 2002 (–5.36 and –5.87 mmꞏyear–1, re-spectively). From 1901 to 1956, annual and monsoon rainfall increased at rates of 3.01 mmꞏyear–1 and 2.98 mmꞏyear–1, respectively. The highest percent reduc-tion was observed in the monsoon season (–21.84%) which had an impact on the decreased annual rainfall (–18.92%) during 1901–2002.

Table 3 contains Z values and mean percent change for annual, monsoon, pre-monsoon, post-mon-soon and winter rainfall at each station in the study area. All 37 stations had significant negative trends on an annual basis with Z values varying from –1.97 to –3.48. The spatial distribution of annual and seasonal rainfall trends at the 37 stations is depicted in Figure 5. Stars and pentagons denote significant and non- -significant negative trends while triangles and dia-monds denote significant and non-significant positive trends, respectively. The maximum reduction in an-nual rainfall was 19–22% in the northwest and south-

Fig. 4. Distribution of average monsoon rainfall (%) with respect to average annual rainfall (1901–2002) at different stations; source: own study

Table 2. Mann–Kendall test and percent change of annual and seasonal rainfall during 1901–2002, 1901–1956 and 1957–2002 for the entire study area

Time 1901–2002 1901–1956 1957–2002

Q Z value change (%) Q Z value change (%) Q Z value change (%) Annual –2.17 –2.95 –18.92 3.01 2.16 13.75 –5.36 –2.23 –23.32 Pre-monsoon 0.05 0.38 8.30 –0.10 –0.30 –6.52 0.68 1.64 43.75 Monsoon –2.12 –3.28 –21.84 2.98 2.29 15.91 –5.87 –2.74 –30.18 Post-monsoon 0.11 0.56 14.14 0.41 1.07 28.28 0.09 0.13 6.00 Winter 0.01 –0.05 3.57 0.18 0.78 24.89 0.02 0.13 3.54

Explanations: Q = magnitude of the trend, Z value = direction; bold value shows significant trend. Source: own study.

Rai

nfal

l, %



Table 3. Mann–Kendall Z values (direction) and change percentage (1901–2002)

Sta-tion No.

Stations Annual Monsoon Pre-monsoon Post-monsoon Winter

Z value

change %

Z value

change %

Z value

change %

Z value

change %

Z value

change %

1 Araria –2.33 –13.39 –3.00 –18.29 0.51 7.05 0.29 5.64 0.56 10.22 2 Aurangabad –2.26 –16.55 –2.19 –18.29 0.06 1.24 0.50 11.20 –0.04 –0.54 3 Banka –3.21 –21.17 –3.72 –24.88 0.31 5.80 1.86 30.65 –1.49 –31.97 4 Begusarai –3.20 –20.01 –3.78 –23.83 0.43 7.90 1.07 20.75 –0.60 –11.53 5 Bhagalpur –3.41 –19.64 –3.79 –24.04 0.40 6.32 1.57 28.01 –0.67 –14.23 6 Bojpur –3.07 –18.47 –2.97 –19.41 0.25 5.48 0.28 6.16 –0.02 –0.13 7 Buxar –2.90 –17.60 –2.76 –18.16 0.32 6.81 0.46 2.41 0.23 3.66 8 Darbhanga –3.23 –20.80 –3.67 –24.31 0.57 10.00 0.37 6.37 0.12 2.75 9 Gaya –2.41 –16.77 –2.37 –19.12 0.10 1.96 0.78 16.62 –0.15 –1.91 10 Gopalganj –3.07 –20.35 –3.19 –22.01 0.25 10.44 –0.29 –5.64 0.68 13.14 11 Jamui –3.22 –20.84 –3.44 –23.20 0.36 5.29 1.43 32.99 –1.16 –21.29 12 Jehanabad –2.89 –18.29 –2.85 –20.03 0.15 3.26 0.46 10.59 –0.01 –0.01 13 Bhabua –2.07 –14.00 –2.12 –14.79 0.19 3.63 0.43 10.10 0.33 6.48 14 Katihar –2.67 –13.62 –3.50 –18.12 0.57 8.04 1.34 22.22 –0.35 –7.65 15 Khagaria –3.15 –20.04 –4.00 –24.14 0.51 8.31 1.06 20.43 –0.71 –13.24 16 Kishanganj –1.97 –10.30 –2.53 –14.33 0.73 9.41 0.48 8.63 0.50 9.04 17 Lakhisarai –3.24 –20.86 –3.64 –23.90 0.41 6.78 1.28 26.20 –0.83 –15.22 18 Madhepura –3.08 –17.37 –3.89 –21.71 0.46 6.55 0.86 14.94 0.04 –0.68 19 Madhubani –3.04 –18.63 –3.43 –22.79 0.53 8.96 0.18 3.65 0.47 10.38 20 Munger –3.24 –20.73 –3.80 –24.90 0.46 7.38 1.30 26.71 –1.05 –22.05 21 Muzafferpur –3.27 –21.72 –3.54 –22.69 0.60 9.53 0.10 2.14 0.10 1.75 22 Nalanda –2.89 –20.17 –3.24 –22.34 0.13 3.01 0.63 14.64 –0.08 –1.22 23 Nawada –2.76 –19.13 –3.02 –21.77 0.17 3.73 1.15 24.72 –0.33 –6.07 24 Paschim Champaran –3.17 –20.59 –3.45 –23.15 0.45 6.96 –1.00 –20.81 1.06 20.54 25 Patna –3.08 –20.08 –3.34 –21.44 –0.19 4.13 0.50 9.88 –0.19 –2.85 26 Purba Champaran –3.08 –20.00 –3.27 –22.07 0.56 8.69 –0.48 –9.10 0.55 13.30 27 Purnia –2.57 –14.08 –3.48 –18.29 0.57 7.81 1.00 16.01 –0.12 –1.83 28 Rohtas –2.25 –15.02 –2.20 –15.97 0.11 2.94 0.39 9.45 0.21 3.58 29 Saharsa –3.18 –18.75 –3.87 –23.35 0.46 7.42 0.90 16.05 –0.23 –6.55 30 Samastipur –3.29 –20.58 –3.77 –25.40 0.54 8.49 0.57 12.59 –0.27 –4.55 31 Saran –3.48 –21.27 –3.26 –22.15 0.47 9.02 0.14 3.85 –0.04 –0.51 32 Sheikhpura –3.09 –19.85 –3.36 –23.21 0.17 3.90 0.94 20.23 –0.30 –5.62 33 Sheohar –3.07 –20.61 –3.39 –21.88 0.54 7.31 –0.15 –1.76 0.31 6.21 34 Sitamarhi –2.98 –19.74 –3.27 –22.54 0.58 8.82 –0.17 –3.62 0.53 10.88 35 Siwan –3.27 –20.96 –3.19 –22.62 0.61 10.28 –0.11 –1.92 0.52 9.81 36 Supaul –2.74 –15.52 –3.46 –21.25 0.43 6.09 0.26 5.64 0.69 13.44 37 Vaishali –3.48 –21.10 –3.55 –23.47 0.44 8.50 0.41 7.86 –0.27 –4.62

<–15% 0–(–15%) 0–15 >15%

Explanation: bold values indicate significant trend. Source: own study.

east parts of the state and the lowest reduction in rain-fall was in the eastern and south-western parts. All 37 stations had more than a 10% decrease in rainfall dur-ing 1901–2002.

The monsoon season showed a dominant and sig-nificant decreasing rainfall trend for all 37 stations in Bihar with Z values ranging from –4.0 to –2.12 (Tab. 3). The maximum drop in monsoon rainfall varied from 23% to 26% in the central part of Bihar although all stations saw a significant decreasing trend ranging from 11% to 26% (Fig. 5). The pre-monsoon season showed non-significant increasing trends at all sta-tions except Patna in central Bihar and Z values ranged from –0.19 to 0.73. The percent change in rainfall was positive over the entire area (1.5–9.0%) with the north-western and eastern parts showing a greater increase than the South-West. The post- -monsoon season showed insignificant mixed trends

and Z values ranged from –1.0 to 1.86. There was a non-significant negative trend in the North-West part of the state ranging from a decrease in rainfall of up –20% to –10%. The greatest drop occurred at the Paschim Champaran station with declines being lower towards the Purba Champaran, Gopalganj, Sheohar, Sitamarhi, Siwan and Buxar stations (up –10% to 0%). Non-significant positive trends were observed at 32 stations in the South-East part of Bihar with the greatest rainfall increase occurring at Banka, Jamui, Lakhisarai and Munger (20–30%).

The winter season had insignificant and mixed trends with the Z value varying from –1.49 to +1.06. The percent change in rainfall trends went from large and positive in the North-West (Paschim Champaran, Purba Champaran, Gopalganj, Sitamarhi, Supul and Madhubani) to large and negative in the South (Banka and Jamui).



Fig. 5. Spatial distribution of annual and seasonal percent change of rainfall and level of significance at 37 stations from 1901 to 2002; stars and pentagons de-note significant and non-significant negative trends, trian-gles and diamonds denote significant and non-significant positive trends, respectively; source: own study

For the annual and monsoon periods of precipita-

tion, the magnitude of trend decreased for all stations. This result is in agreement with CHANDNIHA et al. [2016], who reported a decreasing rate of monsoon rainfall over Jharkhand, which is a neigh boring state of Bihar.

The magnitude of the rainfall trend over 102 years was obtained from Sen’s slope estimator. Figu- re 6 illustrates the change in rainfall rate at each sta-tion in mmꞏyear–1 over the span 1901–2002. The an-nual and monsoon periods clearly showed a reduced rainfall rate while most of the stations in the pre-monsoon, post-monsoon and winter had a rising rate

of change. The maximum decrease is –2.62 mmꞏyear–1 and –2.69 mmꞏyear–1 (Paschim Champaran) and the smallest decrease is –1.34 mmꞏyear–1 and –1.45 mmꞏyear–1 (Kaithar) for monsoon and annual rainfall, respectively. These results reflected the rainfall sce-nario in Bihar between 1901 and 2002 and indicated some noticeable changes. Significant variation in monsoon rainfall has considerable influence on the Indian economy and agriculture. In Bihar, the ob-served decrease in monsoon rainfall may create diffi-culties in the agricultural sectors, industrial and do-mestic water supply, as demand is increasing continu-ously with growing urbanization.



Fig. 6. Change in rainfall rate (mmꞏyear–1) at each station during 1901–2002; source: own study

Fig. 7. Spatial distribution of annual and monsoon rainfall in terms of percentage change from 1901 to 1956; source: own study

Trends from 1901 to 1956

Spatial variation of annual and monsoon rainfall in terms of percent change before the change point (1901–1956) is depicted in Figure 7. Triangles and diamonds represent positive significant and positive non-significant trends, respectively, for each station in the study area. The change in annual and monsoon rainfall is similar across the study region. Z values vary from 1.19 to 2.79, from 1.55 to 2.79, from –1.44 to 0.62, from –0.2 to 2.48 and from –0.08 to 1.52 for annual, monsoon, pre-monsoon, post-monsoon and winter, respectively (Tab. 4). Most of the stations (25 and 29 stations in annual and monsoon periods, re-spectively) were positive significant and some (11 and 8 stations in annual and monsoon periods, respective-ly) were positive non-significant. The range in percent change of rainfall was 2% to 16% and 0% to 25% for annual and monsoon rainfall, respectively. The central part of Bihar had the highest percentage change, 13% to 16% for annual rainfall and 20% to 25% for mon-soon rainfall. The lowest change was noticed at Paschim Champaran, Kishanganj and Bhabua (2% to 5% and 0% to 5% for annual and monsoon rainfall, respectively). The annual and monsoon rainfall of Bihar state significantly increased before the change point of 1956.

Trends from 1957 to 2002

The results of the time series 1957–2002 por-trayed the opposite rainfall scenario to the 1901–1956 period, although it is similar to the entire time series, 1901–2002. Spatial variation of annual and monsoon rainfall in terms of mean percentage change is shown in Figure 8. Stars and pentagons represent the nega-tive significant and negative non-significant trends, respectively, for each station. Most of the stations (29 and 33 stations in annual and monsoon, respectively) were negative significant and a few (8 and 4 stations in annual and monsoon, respectively) were negative non-significant. In the other series (1901–2002 and 1901–1956) there was no significant trend in the pre-monsoon season, but in the latest period (1957–2002) the rainfall changed drastically with 14 stations hav-ing significant increasing trends in Bihar. Z values varied from –3.14 to –0.13, from –3.58 to –0.42, from 0.04 to 2.45, from –2.41 to 1.53 and from –0.8 to 1.7 for annual, monsoon, pre-monsoon, post-monsoon and winter, respectively (Tab. 5). The rate of decline was highest (Gopalganj) in the western part of the state and lowest (Kishanganj) in the eastern part (for 1957–2002). The percent change of rainfall decline was from –31% to –1% and from –40% to –5% for annual and monsoon rainfall, respectively. There was

Rat

e of

rai

nfal

l, m

m∙y

–1




Station No.


Z value

change %

Z value

change %

Z value

change %

Z value

change %

Z value

change %

1 Araria 2.51 12.67 2.45 14.71 –0.93 –18.30 1.38 28.72 0.84 21.80 2 Aurangabad 1.44 13.57 1.75 15.76 0.06 0.63 0.76 15.23 0.40 10.70 3 Banka 2.07 13.63 2.27 15.00 –0.16 –5.57 1.73 45.59 0.12 4.72 4 Begusarai 2.64 16.19 2.75 17.45 –0.18 –4.48 1.46 41.85 0.62 17.78 5 Bhagalpur 2.79 13.21 2.48 16.09 –0.53 –8.68 2.48 39.26 0.83 39.26 6 Bojpur 1.79 13.22 2.11 15.64 –0.08 –1.16 0.82 19.75 0.80 24.78 7 Buxar 1.94 13.28 2.21 15.50 –0.23 –8.59 –0.20 –0.07 1.00 27.27 8 Darbhanga 2.57 15.31 2.65 18.41 –0.29 –6.53 1.17 30.68 1.10 35.96 9 Gaya 1.52 13.58 1.60 14.81 0.09 5.02 0.87 24.21 0.12 4.50 10 Gopalganj 2.17 12.60 2.14 15.72 –0.57 –10.30 0.56 12.00 1.51 40.30 11 Jamui 1.93 14.35 2.04 14.26 –0.08 –2.10 0.11 46.73 –0.08 46.73 12 Jehanabad 1.87 14.04 2.09 15.26 0.13 5.10 0.94 24.19 0.59 13.12 13 Bhabua 1.90 14.80 1.94 18.40 –0.18 –5.14 0.42 8.65 0.95 22.44 14 Katihar 2.20 11.36 2.52 13.14 –0.54 –11.00 1.46 33.33 0.67 20.02 15 Khagaria 2.68 14.52 2.72 16.15 –0.30 –6.89 1.52 40.92 0.88 22.61 16 Kishanganj 2.40 11.68 2.54 15.29 –0.76 –14.98 1.44 23.27 0.70 16.84 17 Lakhisarai 2.37 15.28 2.57 17.29 –0.18 –3.61 1.62 44.94 0.43 12.12 18 Madhepura 2.66 14.09 2.74 16.07 –0.59 –13.37 1.45 37.26 0.95 37.26 19 Madhubani 2.38 15.24 2.45 17.71 –0.45 –10.89 1.14 30.95 1.22 37.90 20 Munger 2.45 15.27 2.65 17.00 –0.23 –3.86 1.55 41.79 0.76 20.90 21 Muzafferpur 2.34 15.00 1.55 16.82 –0.08 –1.39 1.00 29.37 1.14 38.81 22 Nalanda 2.04 13.98 2.10 15.18 –0.09 –5.16 0.86 30.19 0.45 10.96 23 Nawada 1.73 11.91 1.86 14.91 –0.04 –1.71 1.27 39.58 –0.02 –1.06 24 Paschim Champaran 1.19 5.79 1.82 12.22 –1.44 –26.23 0.28 8.73 1.11 33.13 25 Patna 2.00 14.07 2.17 15.69 0.62 –0.38 0.90 27.31 0.62 16.79 26 Purba Champaran 1.97 12.25 1.93 14.78 –0.60 –10.61 0.69 15.66 1.28 40.60 27 Purnia 2.26 12.15 2.52 13.39 –0.63 –13.63 1.44 31.60 0.71 18.75 28 Rohtas 1.63 14.08 1.89 17.34 –0.21 –5.27 0.72 16.36 0.71 20.14 29 Saharsa 2.71 15.88 2.79 16.45 –0.47 –10.36 1.46 38.17 0.97 28.48 30 Samastipur 2.56 17.29 2.76 17.75 –0.12 –3.46 1.22 33.22 0.73 20.32 31 Saran 2.06 13.86 2.37 15.93 –0.11 –3.32 0.94 19.44 0.98 36.42 32 Sheikhpura 2.21 15.31 2.41 16.21 –0.16 –3.11 1.15 40.32 0.39 12.23 33 Sheohar 2.24 13.82 2.10 16.27 –0.29 –4.93 0.97 28.29 1.22 36.41 34 Sitamarhi 2.24 13.80 2.76 16.88 –0.42 –6.32 1.03 27.69 1.12 41.08 35 Siwan 2.30 14.43 2.21 17.43 –0.21 –7.05 0.66 14.00 1.52 39.93 36 Supaul 2.48 14.51 2.55 15.99 –0.81 –15.35 1.31 29.22 0.94 25.69 37 Vaishali 1.85 12.85 2.27 15.92 0.11 1.85 0.98 27.97 0.62 25.07

<–15% 0–(–15%) 0–15 >15%


Fig. 8. Spatial distribution of annual and monsoon rainfall in terms of percent change from 1957 to 2002; source: own study




Station No.


Z value

change %

Z value

change %

Z value

change %

Z value

change %

Z value

change %

1 Araria –0.45 –4.24 –1.17 –11.83 2.56 55.44 –0.05 –0.80 1.53 55.58 2 Aurangabad –2.39 –30.37 –2.42 –34.32 0.51 18.11 0.27 12.90 0.13 3.51 3 Banka –1.57 –16.76 –2.46 –22.57 1.52 41.57 0.08 1.57 0.02 0.74 4 Begusarai –2.41 –23.63 –2.94 –32.61 1.82 47.40 –0.02 –1.02 –0.15 –5.07 5 Bhagalpur –0.30 –16.41 –2.41 –22.63 1.59 39.49 –2.41 1.68 0.64 1.68 6 Bojpur –3.09 –30.35 –3.22 –37.22 0.97 35.27 0.31 11.46 0.17 4.48 7 Buxar –3.07 –31.58 –3.20 –40.28 0.70 27.35 0.47 3.19 0.13 5.69 8 Darbhanga –2.52 –24.03 –3.24 –32.85 2.63 72.82 0.00 –0.30 –0.49 –13.85 9 Gaya –2.29 –28.40 –2.35 –30.89 0.78 21.30 0.17 8.69 0.06 0.74 10 Gopalganj –2.90 –32.25 –3.22 –39.73 1.33 36.99 0.34 16.50 –0.38 –11.11 11 Jamui –2.12 –18.94 –2.59 –25.61 1.48 38.10 1.53 4.73 –0.34 4.73 12 Jehanabad –2.80 –28.69 –2.95 –34.93 0.87 30.71 0.21 7.86 0.21 7.47 13 Bhabua –2.44 –30.60 –2.67 –35.10 0.17 6.18 0.39 17.30 0.47 9.55 14 Katihar –0.68 –5.00 –1.69 –13.64 2.16 44.82 0.19 4.62 1.70 52.88 15 Khagaria –2.03 –19.87 –3.09 –29.87 1.91 44.67 0.00 –0.02 –0.13 –2.70 16 Kishanganj –0.13 2.00 –0.42 –4.99 2.61 54.92 0.13 4.62 1.69 63.16 17 Lakhisarai –2.58 –24.62 –2.84 –31.71 1.53 43.77 0.17 5.69 –0.27 –5.80 18 Madhepura –1.78 –16.55 –2.65 –23.31 2.05 48.61 0.08 2.34 0.61 2.34 19 Madhubani –2.06 –20.54 –2.77 –27.77 2.75 75.85 –0.08 –2.68 –0.27 –11.77 20 Munger –2.16 –23.52 –2.86 –29.77 1.61 40.34 0.17 3.67 –0.19 –7.43 21 Muzafferpur –2.94 –28.57 –3.18 –37.93 2.39 61.71 0.21 11.03 –0.70 –28.41 22 Nalanda –2.50 –28.51 –2.82 –32.68 1.21 35.20 –0.02 –0.33 0.10 2.09 23 Nawada –2.16 –25.52 –2.56 –29.81 1.02 32.24 0.10 4.05 –0.10 –3.74 24 Paschim Champaran –2.65 –26.61 –3.18 –31.67 1.89 44.04 0.34 10.38 0.30 9.83 25 Patna –2.78 –28.81 –3.11 –35.04 0.04 40.89 0.11 5.44 0.04 2.06 26 Purba Champaran –2.61 –28.88 –2.94 –36.17 2.21 50.39 0.55 16.08 –0.78 –23.27 27 Purnia –0.76 –6.65 –1.82 –14.94 2.21 48.50 –0.02 –0.48 1.53 52.36 28 Rohtas –2.42 –29.34 –2.65 –33.77 0.30 11.32 0.33 15.59 0.44 11.76 29 Saharsa –2.01 –19.66 –2.90 –27.88 2.05 46.62 0.00 0.49 0.15 6.84 30 Samastipur –2.56 –27.03 –3.37 –33.43 2.14 58.31 –0.19 –8.00 –0.32 –10.15 31 Saran –3.64 –32.11 –3.58 –39.66 1.63 46.52 0.28 19.84 –0.06 –2.83 32 Sheikhpura –2.63 –26.92 –2.69 –31.56 1.27 35.85 –0.08 –4.15 –0.02 –0.77 33 Sheohar –2.82 –29.64 –3.01 –36.44 2.33 59.38 0.34 14.70 –0.80 –27.10 34 Sitamarhi –2.56 –25.82 –3.37 –34.64 2.73 69.94 0.30 13.05 –0.72 –29.88 35 Siwan –3.24 –31.92 –3.26 –40.91 1.27 39.09 0.46 14.36 –0.42 –13.89 36 Supaul –1.21 –12.87 –2.23 –21.19 2.52 61.26 0.02 0.78 0.81 31.13 37 Vaishali –3.11 –29.59 –3.43 –37.53 1.91 53.74 0.15 7.22 0.02 0.27

<–15% 0–(–15%) 0–15 >15%


a large and significant decrease in percent change of monsoon rainfall (from –40% to –33%) observed at Sheohar, Muzafferpur, Vaishali, Patna, Bhojpur, Bha-bua, Saran, Siwan and Gopalganj in the western part of Bihar.

The present study indicated that there was a sig-nificant rainfall reduction observed over the entire study area from 1957 to 2002 during the monsoon season. On other hand, the post monsoon rainfall in-creased significantly at some stations. The increase in rainfall during the post monsoon may reduce the irri-gation requirements in rabi season, but kharif crops may need supplemental water during the monsoon season. This changing pattern of precipitation may influence future water availability especially for agri-cultural purposes. The economy of the study area is highly dependent on agriculture (monsoon rainfall

dependent) and these results may be very useful for water management and sustainable development. Similar to our study, an increasing trend in precipita-tion in the state of Jharakhand, to the South of Bihar, was found during 1901–1949; this was reversed dur-ing the subsequent period (1950–2011) [CHANDNIHA et al. 2016].

EXTREME EVENTS

The number of extreme rainfall events was calcu-lated and different years were categorized on the basis of extreme events (greater than +10% and –10%) in annual and monsoon rainfall. The extreme events were calculated by taking the difference between av-erage rainfall and actual rainfall and then dividing the difference with the average rainfall. The average rain-



Table 6. Extreme events during different time zones

Year 1901–2002 1901–1956 and 1957–2002

Year 1901–2002 1901–1956 and 1957–2002

annual %

monsoon %

annual %

monsoon %

annual %

monsoon %

annual %

monsoon %

1901 –18.62 –18.54 –22.82 –23.50 1952 2.90 6.78 10.20 15.93

1902 –5.96 4.07 –10.82 –2.26 1953 8.16 18.41 15.84 28.55

1903 –17.69 –22.30 –21.94 –27.03 1954 –4.00 2.48 2.81 11.26

1904 3.09 –6.24 –2.23 –11.95 1955 –0.56 1.61 6.50 10.32

1905 10.84 16.94 5.11 9.82 1956 25.50 19.04 34.41 29.24

1905 4.94 –0.04 –0.48 –6.13 1957 –20.91 –13.97 –15.30 –6.60

1907 –9.51 –9.26 –14.18 –14.78 1958 5.25 2.20 12.72 10.96

1908 –16.08 –17.00 –20.41 –22.05 1959 11.12 0.74 19.01 9.38

1909 6.66 16.76 1.15 9.65 1960 –11.45 –3.91 –5.17 4.33

1910 5.10 9.20 –0.33 2.55 1961 –0.06 0.36 7.04 8.96

1911 13.87 13.64 7.99 6.73 1962 –7.74 –0.46 –1.20 8.06

1912 –18.24 –15.64 –22.46 –20.78 1963 10.02 0.46 17.83 9.07

1913 16.70 14.08 10.67 7.13 1964 –0.77 6.77 6.27 15.92

1914 –9.86 –10.10 –14.52 –15.57 1965 –28.27 –26.11 –23.17 –19.77

1915 0.80 –0.36 –4.40 –6.42 1966 –39.90 –40.19 –35.64 –35.07

1916 12.85 9.71 7.02 3.03 1967 –10.25 –4.23 –3.88 3.97

1917 19.07 6.04 12.92 –0.42 1968 –6.70 –6.68 –0.07 1.32

1918 –3.01 7.13 –8.02 0.61 1969 0.76 4.72 7.91 13.69

1919 13.71 7.82 7.84 1.26 1970 9.74 14.36 17.53 24.16

1920 –0.50 5.48 –5.63 –0.94 1971 31.46 26.60 40.79 37.45

1921 –3.29 5.94 –8.28 –0.51 1972 –15.83 –26.63 –9.86 –20.34

1922 34.33 46.45 27.39 37.54 1973 24.20 11.93 33.02 21.53

1923 –2.20 3.48 –7.25 –2.82 1974 3.31 6.21 10.65 15.31

1924 6.52 5.82 1.02 –0.62 1975 –5.17 –5.69 1.56 2.39

1925 4.01 10.62 –1.36 3.88 1976 –5.13 1.80 1.60 10.52

1926 9.06 5.94 3.43 –0.51 1977 5.23 –4.78 12.70 3.38

1927 –2.98 –8.41 –7.99 –13.99 1978 15.97 9.17 24.20 18.53

1928 9.42 3.03 3.77 –3.24 1979 –8.70 –10.94 –2.22 –3.31

1929 7.10 –8.57 1.57 –14.14 1980 –5.47 –0.05 1.24 8.51

1930 –16.75 –14.97 –21.05 –20.14 1981 –13.20 –10.85 –7.04 –3.22

1931 19.76 24.66 13.58 17.07 1982 –12.48 –14.85 –6.27 –7.56

1932 –8.98 –8.45 –13.68 –14.03 1983 –2.13 –8.58 4.82 –0.75

1933 16.45 16.02 10.44 8.95 1984 22.19 29.82 30.86 40.94

1934 9.84 14.46 4.17 7.49 1985 –9.93 –9.92 –3.54 –2.21

1935 2.99 13.92 –2.33 6.98 1986 –5.81 –4.35 0.87 3.85

1936 32.89 30.25 26.02 22.32 1987 34.82 34.61 44.39 46.14

1937 –1.22 –7.97 –6.32 –13.57 1988 –7.44 –6.76 –0.88 1.23

1938 8.36 12.27 2.77 5.44 1989 –1.92 –1.98 5.04 6.42

1939 9.49 14.72 3.84 7.74 1990 1.92 –1.99 9.15 6.41

1940 2.66 5.34 –2.64 –1.08 1991 –8.63 –4.98 –2.14 3.16

1941 2.26 5.33 –3.02 –1.09 1992 –37.96 –38.77 –33.56 –33.53

1942 16.41 24.50 10.40 16.92 1993 –26.80 –29.83 –21.60 –23.81

1943 18.48 21.94 12.36 14.52 1994 –15.17 –25.26 –9.15 –18.86

1944 8.38 5.54 2.78 –0.89 1995 –25.74 –25.08 –20.47 –18.66

1945 –0.30 –10.74 –5.45 –16.18 1996 –15.99 –22.48 –10.03 –15.84

1946 23.42 22.03 17.05 14.60 1997 –16.34 –23.71 –10.41 –17.17

1947 7.00 11.56 1.47 4.76 1998 –11.04 –29.09 –4.73 –23.02

1948 12.53 12.62 6.72 5.76 1999 –18.30 –16.69 –12.50 –9.55

1949 28.94 26.78 22.28 19.06 2000 –34.50 –30.47 –29.86 –24.51

1950 –1.20 5.32 –6.31 –1.09 2001 –30.04 –35.60 –25.07 –30.08

1951 11.37 13.87 5.62 6.94 2002 –21.07 –27.88 –15.46 –21.70

<-10% >+10%

Note: different colours show different periods. Source: own study.



fall was the average of 102 years (1901–2002), 56 years (1901–1956) and 46 years (1957–2002). Years with extreme rainfall events of more than 10% (posi-tive) from 1901 to 2002 are 1901, 1903, 1908, 1912, 1930, 1957, 1960; 1965–1967, 1972, 1981, 1982 and 1992–2002. Overall, 26 out of 102 years experienced an extreme annual event of negative rainfall, i.e., more than a 10% decrease while 24 years experienced an extreme positive event of rainfall of more than 10% change. During the earlier period (1901–1956), only 5 years of negative extreme events were observed compared to 17 years of positive extreme events an-nually. In the most recent period (1957–2002), 19 years out of 46 had negative extreme events and there were no positive extreme events noticed annually (Tab. 6). The period of 10 years from 1992–2002 was the negative extreme events which reflected the sig-nificant reduction of rainfall in recent years. This study indicated that the frequency of negative extreme events was greater in the latest period (1957–2002) compared to the earlier period. Based on the present study, Bihar is likely to become a drought prone area; there were more negative extreme events over the entire period as well as the latest period (1992–2002) having a continuous 10 years of rainfall reduction. Uncertain and erratic distribution of precipitation as well as a lack of state water resources planning is the major limitation to crop growth in the region.

In this study, annual and monsoon rainfall showed significant variability and rainfall decline was most prominent in the recent 10-year period. These results conform with those of KRISHNAKUMAR et al. [2009], KUMAR, JAIN [2011], DUHAN, PANDEY [2013], PRA-

NUTHI et al. [2014] and JAIN et al. [2017]. Rainfall of summer seasons showed insignificant increasing trends [KRISHNAKUMAR et al. 2009]. In the study ar-ea, all the important irrigation practices are based on monsoon rainfall; based on our study monsoon trends have high variability, which will have effects on irri-gation. It can be seen how trend and change point de-tection techniques are very useful for irrigation as well as water resources planning in this area.

CONCLUSIONS

The present study involved the observation of seasonal and annual precipitation trends at 37 weather stations in the state of Bihar, India from 1901 to 2002. The study showed decreasing rainfall trends at all 37 stations in annual and monsoon periods. Annually, the magnitude of the trend varied from –2.69 mmꞏyear–1 in Paschim Champaran to –1.45 mmꞏyear–1 in Bha-bua. During the monsoon season, a negative signifi-cant trend was obtained at all the stations with the largest declines at Paschim Champaran (–2.62 mmꞏyear–1) and the smallest at Bhabua (–1.34 mmꞏyear–1). Seasonal trends indicated a non-signifi-cant rise in pre-monsoon rainfall with a maximum magnitude at Kishanganj (0.160 mmꞏyear–1) and a minimum magnitude at Aurangabad (0.005

mmꞏyear–1). During the post-monsoon season, most of the stations showed positive non-significant trends but some showed negative non-significant trends. The magnitude of the trend varied from a maximum at Jamui (0.277 mmꞏyear–1) to a minimum at Paschim Champaran (–0.15 mmꞏyear–1). The winter season also did not exhibit any significant increasing or de-creasing trends and a non-significant mixed trend was observed. Overall, a declining trend in rainfall magni-tude was found over Bihar from 1901 to 2002.

The most probable year of change in the observed rainfall trends was 1956. The trend from 1901 to 1956 showed increasing rainfall in Bihar while from 1957 to 2002 the trend was negative. The numbers of ex-treme negative rainfall events have increased in fre-quency during the last 46 years and this resulted in the overall 102-year negative rainfall trend. The resultant annual decrease is about –23.32% from 1957 to 2002 and –18.92% from 1901 to 2002 in Bihar. The impact of monsoon rainfall is found to be dominant over the average annual rainfall. The average monsoon rainfall declined up to –21.84% from 1901–2002, and up to –0.18% from 1957–2002. The declining trend in the latest decade has changed the overall scenario of rain-fall in the study area and this rainfall variability may impact future climatic conditions and agricultural practices. Analyses of the precipitation data should be useful for irrigation and agricultural managers and could play an important role in managing water re-sources more effectively and sustainably in the state.

REFERENCES

AKINREMI O.O., MCGINN S.M., CUTFORTH H.W. 2001. Seasonal and spatial patterns of rainfall trends on the Canadian prairies. Journal of Climate. Vol. 14. No. 9 p. 2177–2182.

ARAGHI A., MOUSAVI BAYGI M., ADAMOWSKI J., MA-

LARD J., NALLEY D., HASHEMINIA H. 2015. Using wavelet transforms to estimate surface temperature trends and dominant periodicities in Iran based on grid-ded reanalysis data. Journal of Atmospheric Research. Vol. 11 p. 52–72.

AYALEW D., TESFAYE K., MAMO G., YITAFERU B., BAYU W. 2012. Variability of rainfall and its current trend in Amhara region, Ethiopia. African Journal of Agricultural Research. Vol. 7(10) p. 1475–1486.

BUISHAND T.A. 1982. Some methods for testing the ho-mogeneity of rainfall records. Journal of Hydrology. Vol. 58. Iss. 1–2 p. 11–27.

CHANDNIHA S.K., MESHRAM S.G., ADAMOWSKI J.F., MESHRAM C. 2016. Trend analysis of precipitation in Jharkhand State, India. Theoretical and Applied Clima-tology. Vol. 130. Iss. 1–2 p. 1–14.

CHEN F.W., LIU C.W. 2012. Estimation of the spatial rain-fall distribution using inverse distance weighting (IDW) in the middle of Taiwan. Paddy and Water Environ-ment. Vol. 10(3) p. 209–222.

CHOWDHURY R.K., BEECHAM S. 2010. Australian rain-fall trends and their relation to the southern oscillation index. Hydrological Processes. Vol. 24(4) p. 504–514.



CUNDERLIK J.M., BURN D.H. 2004. Linkages between regional trends in monthly maximum flows and selected climatic variables. Journal of Hydrologic Engineering. Vol. 9(4) p. 246–256.

CZARNECKA M., NIDZGORSKA-LENCEWICZ J. 2012. Wielo-letnia zmienność sezonowych opadów w Polsce [Long-term variation of seasonal precipitation in Poland]. Wo-da-Środowisko-Obszary Wiejskie. T. 12. Z. 38 p. 45–60.

DUGAM S.S., KAKADE S.B. 2003. Indian monsoon varia-bility in relation to regional pressure index. Journal of Earth System Science. Vol. 112. Iss. 4 p. 521–527.

DUHAN D., PANDEY A. 2013. Statistical analysis of long term spatial and temporal trends of precipitation during 1901–2002 at Madhya Pradesh, India. Atmospheric Re-search. Vol. 122 p. 136–149.

GALLANT A.J., HENNESSY K.J., RISBEY J. 2007. Trends in rainfall indices for six Australian regions: 1910–2005. Australian Meteorological Magazine. Vol. 56(4) p. 223–241.

GOSWAMI B. N., VENUGOPAL V., SENGUPTA D., MAD-HUSOODANAN M.S., XAVIE, P.K. 2006. Increasing trend of extreme rain events over India in a warming environment. Science. Vol. 314. Iss. 5804 p. 1442–1445.

GUHATHAKURTA P., RAJEEVAN M. 2008. Trends in the rainfall pattern over India. International Journal of Cli-matology. Vol. 28(11) p. 1453–1470.

HAMMOURI N., AL-QINNA M., SALAHAT M., ADA-

MOWSKI J., PRASHER S.O. 2015. Community based adaptation options for climate change impacts on water resources: The case of Jordan. Journal of Water and Land Development. No. 26 p. 3–17.

HENNESSY K.J. 1999. Australian rainfall changes, 1910–1995. Australian Meteorological Magazine 48 p. 1–13.

INAM A., ALBANO R., ADAMOWSKI J., PRASHER S. 2017. Parameter estimation and uncertainty analysis of the Spatial Agro Hydro Salinity Model (SAHYSMOD) in the semi-arid climate of Rechna Doab, Pakistan. Envi-ronmental Modelling and Software. Vol. 94 p. 186–211.

India Water Portal 2010. Meteorological datasets [online]. [Access 2.05.2017]. Available at: http://www. indiawaterportal.org/metdata

IPCC 2007. Climate change 2007 – The scientific basis. Contribution of Working Group I to the Fourth Assess-ment Report of the Intergovernmental Panel for Climate Change [online]. Intergovernmental Panel for Climate Change. [Access 15.04.2017]. Available at: https://www.ipcc.ch/publications_and_data/publications_ipcc_fourth_assessment_report_wg1_report_the_physical_science_basis.htm

JAGADEESH P., ANUPAMA C. 2014. Statistical and trend analyses of rainfall: A case study of Bharathapuzha River basin, Kerala, India. ISH Journal of Hydraulic Engineering. Vol. 20(2) p. 119–132.

JAIN S.K., NAYAK P.C., SINGH Y., CHANDNIHA S.K. 2017. Trend in rainfall and peak flows for some river basins in India. Current Science. Vol. 112(8) p. 1712–1726.

KARL T.R., EASTERLING D.R. 1999. Climate extremes: se-lected review and future research directions. Climatic Change. Vol. 42. Iss. 1 p. 309–325.

KENDALL M.G. 1975. Rank correlation methods. 4th ed. London. Charles Griffin pp. 202.

KOTHYARI U.C., SINGH V.P. 1996. Rainfall and temperature trends in India. Hydrological Processes. Vol. 10(3) p. 357–372.

KRISHNAKUMAR K.N., RAO G.P., GOPAKUMAR C.S. 2009. Rainfall trends in twentieth century over Kerala, India. Atmospheric Environment. Vol. 43(11) p. 1940–1944.

KUMAR V., JAIN S.K. 2011. Trends in rainfall amount and number of rainy days in river basins of India (1951–2004). Hydrology Research. Vol. 42(4) p. 290–306.

LEBEL T., BASTIN G., OBLED C., CREUTIN J.D. 1987. On the accuracy of areal rainfall estimation: a case study. Wa-ter Resources Research. Vol. 23(11) p. 2123–2134.

LU S., WU B., WEI Y., YAN N., WANG H., GUO S. 2015. Quantifying impacts of climate variability and human activities on the hydrological system of the Haihe River Basin, China. Environmental Earth Sciences. Vol. 73(4) p. 1491–1503.

MARTINO G. D., FONTANA N., MARINI G., SINGH V.P. 2012. Variability and trend in seasonal precipitation in the continental United States. Journal of Hydrologic Engi-neering. Vol. 18(6) p. 630–640.

MOHAPATRA M., MOHANTY U.C., BEHERA S. 2003. Spatial variability of daily rainfall over Orissa, India, during the southwest summer monsoon season. International Jour-nal of Climatology. Vol. 23(15) p. 1867–1887.

NICHOLLS N., LAVERY B. 1992. Australian rainfall trends during the twentieth century. International Journal of Climatology. Vol. 12(2) p. 153–163.

PANDEY R.P., RAMASASTRI K.S. 2001. Relationship be-tween the common climatic parameters and average drought frequency. Hydrological Processes. Vol. 15(6) p. 1019–1032.

PANT G. B., HINGANE L.S. 1988. Climatic changes in and around the Rajasthan desert during the 20th century. Journal of Climatology. Vol. 8(4) p. 391–401.

PARTAL T., KAHYA E. 2006. Trend analysis in Turkish pre-cipitation data. Hydrological Processes. Vol. 20(9) p.2011–2026.

PARTHASARATHY B., DHAR O.N. 1974. Secular variations of regional rainfall over India. Quarterly Journal of the Roy-al Meteorological Society. Vol. 100(424) p. 245–257.

PARTHASARATHY B., DHAR O.N. 1978. Climate fluctuations over Indian Region–Rainfall: A Review. Vol. 31. Indian Institute of Tropical Meteorology. Pune Research Re-port No. RR-025.

PATTANAIK D., RAJEEVAN M. 2010.Variability of extreme rainfall events over India during southwest monsoon season. Meteorological Applications. Vol. 17 p. 88–104.

PATTANAIK D.R., RAJEEVAN M. 2007. Northwest Pacific tropical cyclone activity and July rainfall over India. Meteorology and Atmospheric Physics. Vol. 95(1–2) p. 63–72.

PETRIE M.D., COLLINS S.L., GUTZLER D.S., MOORE D.M. 2014. Regional trends and local variability in monsoon precipitation in the northern Chihuahuan Desert, USA. Journal of Arid Environments. Vol. 103 p. 63–70.

PHILANDRAS C.M., NASTOS P.T., KAPSOMENAKIS J., DOUVIS K.C., TSELIOUDIS G., ZEREFOS C.S. 2011. Long term precipitation trends and variability within the Mediter-ranean region. Natural Hazards and Earth System Sci-ences. Vol. 11(12) p. 3235–3250.

PINGALE S., ADAMOWSKI J., JAT M., KHARE D. 2015. Impli-cations of spatial scale on climate change assessments. Journal of Water and Land Development. No. 26 p. 37–55.

PINGALE S.M., KHARE D., JAT M.K., ADAMOWSKI J. 2016. Trend analysis of climatic variables in an arid and semi-arid region of the Ajmer District, Rajasthan, India. Jour-nal of Water and Land Development. No 28 p. 3–18.



PRANUTHI G., DUBEY S.K., TRIPATHI S.K., CHANDNIHA S.K. 2014. Trend and change point detection of precipitation in urbanizing Districts of Uttarakhand in India. Indian Journal of Science and Technology. Vol. 7(10) p. 1573–1582.

RODRIGUEZ-PUEBLA C., ENCINAS A.H., NIETO S., GARMEN-DIA J. 1998. Spatial and temporal patterns of annual precipitation variability over the Iberian Peninsula. In-ternational Journal of Climatology. Vol. 18(3) p. 299–316.

SARKAR S., SINGH R.P., KAFATOS M. 2004. Further evi-dences for the weakening relationship of Indian rainfall and ENSO over India. Geophysical Research Letters. Vol. 31. Iss. 13. DOI 10.1029/2004GL020259.

SEN P.K. 1968. Estimates of the regression coefficient based on Kendall's tau. Journal of the American Statistical As-sociation. Vol. 63(324) p. 1379–1389.

SEN ROY S., BALLING R. 2004. Trends in extreme daily precipitation indices in India. International Journal of Climatology. Vol. 24 p. 457–466.

SINGH V.P., CHOWDHURY P.K. 1986. Comparing some methods of estimating mean areal rainfall. Journal of the American Water Resources Association. Vol. 22(2) p. 275–282.

SOLOMON S., QIN D., MANNING M., ALLEY R.B., BERNTSEN T., BINDOFF N.L., CHEN Z., CHIDTHAISONG A., GREGO-

RY J.M., HEGERL G.C., HEIMANN M., HEWITSON B., HOSKINS B.J., JOOS F., JOUZEL J., KATTSOV V., LOHMANN U., MATSUNO T., MOLINA M., NICHOLLS N., OVERPECK J., RAGA G., RAMASWAMY V., REN J., RUS-

TICUCCI M., SOMERVILLE R., STOCKER T.F., WHETTON P., WOOD R.A., WRATT D. 2007. Technical summary. In: Climate change 2007: The physical science basis. Contribution of Working Group I to the Fourth Assess-ment Report of the Intergovernmental Panel on Climate Change. Eds. S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor, H.L. Mil-

ler. Cambridge, United Kingdom, New York, NY Cam-bridge University Press, USA p. 19–91.

TABARI H., TALAEE P.H., EZANI A., SOME’E B.S. 2012. Shift changes and monotonic trends in autocorrelated temperature series over Iran. Theoretical and Applied Climatology. Vol. 109. Iss. 1–2 p. 95–108.

TASCHETTO A.S., ENGLAND M.H. 2009. An analysis of late twentieth century trends in Australian rainfall. Interna-tional Journal of Climatology. Vol. 29(6) p. 791–807.

TEAM C.W., PACHAURI R.K., MEYER L.A. 2014. IPCC, 2014: Climate change 2014: Synthesis report. Contribu-tion of Working Groups I, II and III to the Fifth As-sessment Report of the intergovernmental panel on Cli-mate Change. Geneva, Switzerland pp. 151.

THEIL H. 1950a. A rank-invariant method of linear and pol-ynomial regression analysis. P. 1. Proceedings of the Royal Netherlands Academy of Sciences. No. 53 p. 386–392.

THEIL H. 1950b. A rank-invariant method of linear and pol-ynomial regression analysis. P. 2. Proceedings of the Royal Netherlands Academy of Sciences. No. 53 p. 521–525.

WARWADE P., SHARMA N., PANDEY A., AHRENS B. 2016. Analysis of climate variability in a part of Brahmaputra river basin in India. In: River system analysis and man-agement. Ed. N. Sharma. Singapore. Springer p. 113–142.

YUE S., HASHINO M. 2003. Long term trends of annual and monthly precipitation in Japan. Journal of the American Water Resources Association. Vol. 39(3) p. 587–596.

ŻMUDZKA E. 2012. Wieloletnie zmiany zasobów termicz-nych w okresie wegetacyjnym i aktywnego wzrostu roślin w Polsce [Long-term changes in thermal re-sources during vegetation and active growth of plants in Poland]. Woda-Środowisko-Obszary Wiejskie. T. 12. Z. 2(38) p. 377–389.

Pratibha WARWADE, Shalini TIWARI, Sunil RANJAN, Surendra K. CHANDNIHA, Jan ADAMOWSKI

Przestrzenna i czasowa zmienność opadów w stanie Bihar w Indiach

STRESZCZENIE

W badaniach prezentowanych w niniejszej pracy wykryto po raz pierwszy długookresowe trendy rocznych i sezonowych wartości opadów w indyjskim stanie Bihar w latach 1901–2002. Stosując kumulatywny test od-chyleń (CUSUM – ang. cumulative deviation test) i regresję liniową zidentyfikowano punkt zwrotny. Następnie serie czasowe zostały podzielone na dwie grupy: przed i po punkcie zwrotnym. Do oceny przestrzennego roz-kładu trendów zastosowano program Arc-Map 10.3. Stwierdzono, że trendy rocznych i monsunowych opadów znacząco malały. Nie zaobserwowano istotnych trendów w odniesieniu do opadów przed monsunem, po monsu-nie i w okresie zimowym. Średnie zmniejszenie ilości opadów wynosiło 2,17 mm rok–1 i 2,13 mm rok–1 odpo-wiednio dla opadów rocznych i monsunowych. Prawdopodobnym punktem zwrotnym był rok 1956. Liczba skrajnych negatywnych zjawisk była większa w okresie 1957–2002 niż w okresie 1901–1956. Słowa kluczowe: hydrologia, opady, zmiana klimatu, zmienność w czasie i przestrzeni

reviewed 17.10.2017 accepted 08.11.2017 a spatio-temporal

Documents